Why I Built llamadart: Offline Local LLM Inference for Dart/Flutter

I built a desktop AI-powered writing assistant, and cloud inference with Gemini worked great. But I...